Web Analytics 101: Think Different
ByAs much as some people may think that web analytics has some sort of mystical or voodoo powers and that by using them you will be filled with a wealth of insight and knowledge that you will immediately see working its wonders on your site’s traffic, think again. All web analytic software does is give you a bunch of dumb data. You (or who ever is in charge of the analytics) play the pivotal role of what information you need and how to use it.
In order to properly interpret data you must first be able to ask the right questions. Merely saying, I want to know where my customers come from and what they do is not a specific enough question.
When you start formulating the focus of your analytical endeavors it must be customer centric. You must think to yourself, what is the primary purpose of my site (or why is the customer visiting me?). You need to think about what tasks the customer must perform in order for them to meet the purpose of the site. You need to think about how you can achieve maximum customer satisfaction and what needs need to be filled in order to get customer satisfaction. Last question would be how can you improve your customers’ experience?
Once you have figured these out then you look at what metrics apply to these questions. For instance, bounce and conversion rates, revenue streams, keyword and link referrers, etc.
Metrics should not stop with simply getting a Web analytics tool. Web analytics will tell you that customers leave your purchase page and that 20% of them buy, but it won’t tell you why they leave. This is were studies and surveys help. They will fill in the whys and hows while analytic software tells you the whats and wheres.
The maintenance portion of you web analytics should be kept up by an expert. As Avinash Kaushik stated in his book Web Analytics An Hour a Day (a great must read by the way, get a copy), you must apply the 90/10 rule. For every ten bucks you spend on tools you should spend 90 on staff. Quality information can be achieved by quality staff. If you are intent on doing this yourself (which is a big mistake for mid size companies and larger) then you should spend $90 on getting yourself educated. You will only be able to take the analysis so far (there are on 24 hours in a day), but the knowledge gained will help you leverage the tools you have and will give you a better understanding of the job when you decide to hirer an analyst.
Again I greatly recommend getting Web Analytics An Hour a Day. It is roughly 400 pages of concrete information you can use to better explore this often neglected area of untapped potential. Until the next time, happy hosting!